with high throughput. [21][22][23] By interrupting the fluid flow during the exposure through an interference phase mask, the stop-flow lithography technique enables high-resolution fabrication of patterned particles. [24][25][26] In combination with digital mirror devices, the vertical-flow lithography (VFL) method can synthesize anisotropic particles in real time using a light source that is aligned with the fluid flow. [27,28] To overcome the limiting resolution of these projection techniques, scanning two-photon lithography is coupled with CFL for the synthesis of truly 3D particles and fibers. [29,30] The current two-photon CFL (TP-CFL) methods offer fast in-plane laser writing speeds surpassing 1 mm s −1 using scanning galvanometer mirrors; however, the out-of-plane writing speed is still limited by the piezo-driven nanostage movement of typically 100 µm s −1 . [31] The transfer of the continuous real-time synthesis from particles (0D), fibers (1D), and planar geometries (2D) toward microtubes (3D) represents an intricate challenge with respect to the synthesis of i) high aspect ratio microtubes with ii) rigorous control over the morphology and surface topology by coupling scanning two-photon lithography with a vertical flow to obtain iii) high in-plane and out-of-plane synthesis speeds with iv) sub-micrometer resolution. Below, for the first time, we demonstrate two-photon VFL (TP-VFL) that combines the advantages of TP-CFL with VFL to control the out-of-plane fabrication speed by the vertical flow inside a microfluidic channel. This unique approach is an enabling step toward the synthesis of tubular scaffolds for vascular tissue engineering, microstents for intravascular pressure reduction, microneedles for transdermal drug delivery, nerve guides for neuronal regeneration, and porous hollow fiber membranes for separation applications. [32][33][34][35][36] Results and Discussion Fluid Flow-Coupled Two-Photon PolymerizationTo validate the proposed hypothesis, we design and fabricate a microfluidic chip featuring precise fluid-flow control (Figure 1). The designed master mold (Figure 1a) is printed onto a glass slide using maskless dip-in laser lithography (Figure 1b). [37,38] Soft-lithography replica molding is used to obtain the silicone microfluidic chip plasma bonded to a glass slide (Figure 1c). [39] Two-photon vertical-flow lithography is demonstrated for synthesis of complex-shaped polymeric microtubes with a high aspect ratio (>100:1). This unique microfluidic approach provides rigorous control over the morphology and surface topology to generate thin-walled (<1 µm) microtubes with a tunable diameter (1-400 µm) and pore size (1-20 µm). The interplay between fluid-flow control and two-photon lithography presents a generic high-resolution method that will substantially contribute toward the future development of biocompatible scaffolds, stents, needles, nerve guides, membranes, and beyond. Lithography

Due to its biocompatibility, electrical conductivity, and tissue‐like elasticity, poly(3,4‐ethylenedioxythiophene):polystyrene sulfonate (PEDOT:PSS) constitutes a highly promising material regarding the fabrication of smart cell culture substrates. However, until now, high‐throughput synthesis of pure PEDOT:PSS geometries was restricted to flat sheets and fibers. In this publication, the first microfluidic process for the synthesis of spherical, highly porous, pure PEDOT:PSS particles of adjustable material properties is presented. The particles are synthesized by the generation of PEDOT:PSS emulsion droplets within a 1‐octanol continuous phase and their subsequent coagulation and partial crystallization in an isopropanol (IPA)/sulfuric acid (SA) bath. The process allows to tailor central particle characteristics such as crystallinity, particle diameter, pore size as well as electrochemical and mechanical properties by simply adjusting the IPA:SA ratio during droplet coagulation. To demonstrate the applicability of PEDOT:PSS particles as potential cell culture substrate, cultivations of L929 mouse fibroblast cells and MRC‐5 human fibroblast cells are conducted. Both cell lines present exponential growth on PEDOT:PSS particles and reach confluency with cell viabilities above 95 vol.% on culture day 9. Single cell analysis could moreover reveal that mechanotransduction and cell infiltration can be controlled by the adjustment of particle crystallinity.

Organic neuromorphic device networks can accelerate neural network algorithms and directly integrate with microfluidic systems or living tissues. Proposed devices based on the bio-compatible conductive polymer PEDOT:PSS have shown high switching speeds and low energy demand. However, as electrochemical systems, they are prone to self-discharge through parasitic electrochemical reactions. Therefore, the network's synapses forget their trained conductance states over time. This work integrates single-device high-resolution charge transport models to simulate entire neuromorphic device networks and analyze the impact of self-discharge on network performance. Simulation of a single-layer nine-pixel image classification network commonly used in experimental demonstrations reveals no significant impact of self-discharge on training efficiency. And, even though the network's weights drift significantly during self-discharge, its predictions remain 100% accurate for over ten hours. On the other hand, a multi-layer network for the approximation of the circle function is shown to degrade significantly over twenty minutes with a final mean-squared-error loss of 0.4. We propose to counter the effect by periodically reminding the network based on a map between a synapse's current state, the time since the last reminder, and the weight drift. We show that this method with a map obtained through validated simulations can reduce the effective loss to below 0.1 even with worst-case assumptions. Finally, while the training of this network is affected by self-discharge, a good classification is still obtained. Electrochemical organic neuromorphic devices have not been integrated into larger device networks. This work predicts their behavior under nonideal conditions, mitigates the worst-case effects of parasitic self-discharge, and opens the path toward implementing fast and efficient neural networks on organic neuromorphic hardware.

No abstract